PPT-podklad-modra Predikce proteinové struktury PPT-podklad-modra LL_logo_neg_RGB PPT-podklad-modra qPredikce sekundární struktury qPredikce proteinového foldu qPredikce terciární struktury qPredikce molekulárních komplexů qHodnocení predikčních metod q q § Osnova 2/39 Predikce proteinové struktury PPT-podklad-modra qPredikce sekundární struktury qPredikce proteinového foldu §Navlékání - angl. Threading qPredikce terciární struktury §Homologní modelování - angl. Homology modelling §Ab initio predikce - angl. Ab initio prediction qPredikce molekulárních komplexů §Molekulární dokování - angl. Molecular docking Predikce proteinové struktury 3/39 Predikce proteinové struktury PPT-podklad-modra qPřiřazení jednoho konformačního stavu každému aminokyselinovému zbytku v proteinové sekvenci: §a-šroubovice (H, angl. helix) §b-řetězec (E, angl. strand) §otočka (C, angl. coil) Predikce sekundární struktury 4/39 Predikce proteinové struktury PPT-podklad-modra qPřiřazení jednoho konformačního stavu každému aminokyselinovému zbytku v proteinové sekvenci: §Přesnost >80% §Klasifikace proteinů §Identifikace proteinových domén a funkčních motivů §Zlepšení spolehlivosti sekvenčních přiložení §Příprava na predikci terciární struktury Predikce sekundární struktury 5/39 Predikce proteinové struktury PPT-podklad-modra Predikce sekundární struktury C:\Dokumenty\VYUKA\Bioinformatics\SCAN\11_3_fig_baxevanis.bmp 6/39 Predikce proteinové struktury PPT-podklad-modra qPSI-PRED §Kombinuje evoluční informaci s predikcí neuronovou sítí Predikce sekundární struktury 7/39 Predikce proteinové struktury PPT-podklad-modra qQuick2D §Přiřazení sekundárních elementů: α-šroubovic, β-řetězců, otoček, transmembránových šroubovic a neuspořádaných regionů §Metody PSI-PRED, JNET, Prof, Coils, MEMSAT2, HMMTOP, ... Predikce sekundární struktury 8/39 Predikce proteinové struktury PPT-podklad-modra qGeneSilico MetaServer §Meta-server pro predikci struktury proteinů, včetně predikce sekundární elementů = konsensus Predikce sekundární struktury 9/39 Predikce proteinové struktury PPT-podklad-modra q q q qNavlékání §Rozpoznávání proteinového foldu §Hledá strukturu, která nejlépe odpovídá proteinové sekvenci prohledáváním knihovny známých foldů a hodnocením skóre §Používá se pro struktury, pro které není k dispozici vhodný templát pro homologní modelování §Neposkytne výsledek, pokud správný fold není v knihovně Predikce proteinového foldu 10/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra MSLGAKPFGE... q q q modelovaná sekvence qNavlékání Predikce proteinového foldu 11/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q fold 1 fold 2 fold n qNavlékání q q Predikce proteinového foldu MSLGAKPFGE... modelovaná sekvence 12/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q konstrukce modelu qNavlékání q q Predikce proteinového foldu fold 1 fold 2 fold n MSLGAKPFGE... modelovaná sekvence 13/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q qNavlékání q q Predikce proteinového foldu konstrukce modelu fold 1 fold 2 fold n MSLGAKPFGE... modelovaná sekvence výpočet energie 14/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q výpočet skóre a klasifikace qNavlékání q q Predikce proteinového foldu konstrukce modelu fold 1 fold 2 fold n MSLGAKPFGE... modelovaná sekvence výpočet energie 15/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q qNavlékání §PHYRE §GenTHREADER § Predikce proteinového foldu 16/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q Predikce terciární struktury qHomologní modelování qAb initio predikce § § 17/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q qHomologní modelování §Vytváří atomistický model založený na experimentálně určené struktuře, která je sekvenčně blízce příbuzná §Vyžadovaná sekvenční identita >25% §Základní princip = struktura je konzervována déle než sekvence Predikce terciární struktury 18/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q qHomologní modelování Predikce terciární struktury MSLGAKPFGE... modelovaná sekvence 19/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q prohledání databáze párovým přiložením qHomologní modelování Predikce terciární struktury MSLGAKPFGE... modelovaná sekvence 20/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q identifikace templátu qHomologní modelování Predikce terciární struktury prohledání databáze párovým přiložením MSLGAKPFGE... modelovaná sekvence 21/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra přiložení sekvencí q q q MSLGAKPFGE... MGV-AKTYGE... qHomologní modelování Predikce terciární struktury identifikace templátu prohledání databáze párovým přiložením MSLGAKPFGE... modelovaná sekvence 22/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q extrakce páteře náhrada vedl. řetězců qHomologní modelování Predikce terciární struktury přiložení sekvencí MSLGAKPFGE... MGV-AKTYGE... identifikace templátu prohledání databáze párovým přiložením MSLGAKPFGE... modelovaná sekvence 23/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q doplnění smyček qHomologní modelování Predikce terciární struktury q q q extrakce páteře náhrada vedl. řetězců přiložení sekvencí MSLGAKPFGE... MGV-AKTYGE... identifikace templátu prohledání databáze párovým přiložením MSLGAKPFGE... modelovaná sekvence 24/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q optimalizace modelu qHomologní modelování Predikce terciární struktury doplnění smyček extrakce páteře náhrada vedl. řetězců přiložení sekvencí MSLGAKPFGE... MGV-AKTYGE... identifikace templátu prohledání databáze párovým přiložením MSLGAKPFGE... modelovaná sekvence 25/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q hodnocení modelu qHomologní modelování Predikce terciární struktury q q q optimalizace modelu doplnění smyček extrakce páteře náhrada vedl. řetězců přiložení sekvencí MSLGAKPFGE... MGV-AKTYGE... identifikace templátu prohledání databáze párovým přiložením MSLGAKPFGE... modelovaná sekvence 26/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q qHomologní modelování §Swiss-Model §Modeller q q Predikce terciární struktury 27/39 Predikce proteinové struktury model based on experimentally determined structure that is closely related to target sequence Homology modelling: Suppose a target protein, of known amino acid sequence but unknown structure, is homologous to one or more proteins of known structure. Then we expect that much of the structure of the target protein will resemble that of the known protein, and it can serve as a basis for a model of the target structure. The completeness and quality of the result depend crucially on how similar the sequences are. As a rule of thumb, if the sequences of two related proteins have 50% or more identical residues in an optimal alignment, the proteins are likely to have similar conformations over more than 90% of the structures. (This is a conservative estimate, as the following illustration shows.) PPT-podklad-modra q q q qAb initio predikce §Vytváří atomistický model založený na základních fyzikálních principech §Hledá geometrii struktury v globálním energetickém minimu §Umožňuje navrhovat struktury neexistující v přírodě §“Svatý Grál“ bioinformatiky q q Predikce terciární struktury 28/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q qAb initio predikce §Rosetta, Robetta Predikce terciární struktury 29/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q qMeta-servery §GeneSilico §3D-Jury § q q Predikce terciární struktury 30/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q qMolekulární dokování §Umísťování malých organických molekul – ligandů – do vazebných domén receptorů, aktivních center enzymů nebo žlábků DNA §Náhodně generované orientace a konformace ligandu v blízkosti biomolekuly jsou hodnoceny energetickým skóre §Energetické skóre = interakční energie = van der Waalsova energie + elektrostatická energie + energie vodíkových vazeb + entropie Predikce molekulárních komplexů 31/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q qMolekulární dokování §DOCK §AUTODOCK § q q Predikce molekulárních komplexů 32/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q Hodnocení predikčních metod qCASP §Critical Assessment of Techniques for Protein Structure Prediction §Mezinárodní soutěž spolehlivosti predikčních metod = umožňuje kritické a objektivní hodnocení §K hodnocení jsou využívány slepé predikce = soutěžící obdrží proteinové sekvence se známou, avšak dosud nepublikovanou strukturou – organizátoři porovnají predikované a experimentální struktury 33/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q Hodnocení predikčních metod qCASP §Predikce terciární struktury §Predikce molekulárních komplexů §Predikce kontaktů mezi zbytky §Predikce neuspořádaných regionů §Predikce domén §Predikce funkce proteinů §Hodnocení kvality modelů §Upřesnění modelů 34/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra q q q Hodnocení predikčních metod qCASP 35/39 Predikce proteinové struktury predicts structural fold of target protein sequence by fitting sequence into structural database and selecting the best-fitting fold Fold recognition: Given a library of known protein structures and their amino acid sequences, and the amino acid sequence of a protein of unknown structure, can we find the structure in the library that is most likely to have a folding pattern similar to that of the protein of unknown structure? PPT-podklad-modra Predikce proteinové struktury 36/39 Proteinové inženýrství Bi7410 §Období: jaro §Rozsah: přednáška 1 hodina/týden §Vyučující: Mgr. Radka Chaloupková, Ph.D. §Osnova: §strukturně-funkční vztahy proteinů §metody exprese a purifikace rekombinantních proteinů §metody strukturní a funkční analýzy proteinů §racionální design, semi-racionální design a řízená evoluce §příklady využití proteinového inženýrství § § § § https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcQrLKXrQgnc9DlB7Dji8Sdi_MLonMc71WWImfWJXo-2kui 6PNv7 http://www.rssynthesis.com/wp-content/uploads/2013/01/recombinant_protein.jpg http://climate.nasa.gov/images/NewPowerPlants1.jpg PPT-podklad-modra Predikce proteinové struktury 37/39 Strukturní biologie Bi9410+9410c §Období: podzim §Rozsah: přednáška 2 hodiny/týden, cvičení 2 hodiny/týden §Vyučující: Mgr. David Bednář, Ph.D. §Osnova: §struktura, stabilita a dynamika biologických makromolekul §makromolekulární interakce a komplexy §stanovení a předpověď struktury, identifikace důležitých oblastí §stanovení vlivu mutace na strukturu a funkci proteinu §aplikace v biologickém výzkumu, návrhu léčiv a biokatalyzátorů § § § § § § § https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcQ73isI4CaascsbIrO9P6EUkn81mbBATNitcfnR1H0Z1zj 3y-peNQ dock http://pymol.sourceforge.net/pmimag/800/virus3600.jpg PPT-podklad-modra Predikce proteinové struktury 38/39 Molekulární biotechnologie Bi7430 §Období: podzim §Rozsah: přednáška 2 hodiny/týden, cvičení 2 hodiny/týden §Přednášky: Prof. Zbyněk Prokop, Ph.D. §Cvičení: Dr. Šárka Bidmanová, Dr. Koen Beerens, Dr. Veronika Štěpánková, Mgr. Lukáš Chrást §Osnova: §proteinové, metabolické a tkáňové inženýrství §genetické inženýrství rostlin a živočichů §molekulární diagnostika, vakcíny, terapeutika §buněčná a genová terapie, regenerativní medicína §molekulární biotechnologie v průmyslu a zemědělství (c) 2008 by SM Carr, after Genetix, with permission http://us.123rf.com/400wm/400/400/jarun011/jarun0111109/jarun011110900110/10517468-96-wells-plate-f or-elisa-immunology-testing-method.jpg standard File:Golden Rice.jpg PPT-podklad-modra qClaverie, J-M., & Notredame, C. (2006). Bioinformatics For Dummies (2nd ed.). Wiley Publishing, Hoboken, p. 436. qXiong, J. (2006). Essential Bioinformatics. Cambridge University Press, New York, p. 352. q qPSI-PRED: http://bioinf.cs.ucl.ac.uk/psipred/psiform.html qQuick2D (MPI Toolkit): http://toolkit.tuebingen.mpg.de/quick2_d qModeller: http://salilab.org/modeller/ qModeller (GeneSilico): https://genesilico.pl/toolkit/unimod?method=Modeller qSwiss-Model: http://swissmodel.expasy.org/ qGenTHREADER: http://bioinf.cs.ucl.ac.uk/psipred/psiform.html qPHYRE: http://www.sbg.bio.ic.ac.uk/~phyre/index.cgi qGeneSilico MetaServer: https://www.genesilico.pl/meta2/ q3D-Jury: http://meta.bioinfo.pl/submit_wizard.pl qRosetta@home: http://boinc.bakerlab.org/rosetta/ qCASP: http://predictioncenter.org/index.cgi q q q q q q q q q q q q q q q q q q q q q Reference 39/39 Predikce proteinové struktury